Google DeepMindproprietary
Released: 2026-07-21

Gemini 3.6 Flash

Google high-efficiency workhorse model preceding 3.7 Flash with outstanding throughput and low token costs.

Intelligence Index
91.8/ 100
Calibrated multi-domain composite
Throughput Speed
175tok/sec
Output streaming throughput
Time To First Token
250ms
Initial chunk server latency
LiveBench Contam-Free
77.2%
September 2026 suite score

LiveBench Multi-Domain Evaluation Breakdown

Monthly refreshed contamination-free test sets
Logical Reasoning86.5%
Coding & Repo Repair77.8%
Mathematics (AIME)91.2%
Data Analysis & Tables75.0%
Language & Comprehension70.0%
Instruction Following64.5%

Sourced Benchmark Evaluations (3)

Verified performance across authoritative benchmarks with provenance tracking.
BenchmarkStatusScoreTrustDateSource TypeProvenance
ARC-AGI-2Nearing Saturation71.4%
75
2026-08-18independentSource ↗
SWE-bench VerifiedSaturated59.1%
25
2026-07-30independentSource ↗
IFEvalDeprecated64.5%
25
2026-07-25independentSource ↗
OpenAI Compatible API ExecutionAPI Ready
from openai import OpenAI

client = OpenAI()
response = client.chat.completions.create(
    model="gemini-3-6-flash",
    messages=[{"role": "user", "content": "Evaluate multi-step logic problem"}],
    temperature=0.2
)
print(response.choices[0].message.content)

Evaluation & Sourcing Notes

Scores listed for Gemini 3.6 Flash represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.

All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.